Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computing system, comprising: one or more processors; and a memory storing program instructions that when executed by the one or more processors cause the one or more processors to at least: receive from a user device, a text query; determine that the text query corresponds to a defined category; provide an image refinement option in response to a determination that the text query corresponds to the defined category; determine and return a plurality of results corresponding to the text query; subsequent to receiving the text query, receive from the user device an image of an object as part of the image refinement option; generate an object feature vector representative of the object; compare the object feature vector with a plurality of stored feature vectors corresponding to segments of images returned as the plurality of results to determine a plurality of respective similarity scores, each of the respective similarity scores representative of a similarity between the object feature vector and a respective one of the plurality of stored feature vectors; generate a ranked list of the plurality of results based at least in part on the plurality of respective similarity scores of the object feature vector with the plurality of stored feature vectors; and present the ranked list in response to the receipt of the image.
Image search and retrieval. A system addresses the problem of refining text-based image search results using user-provided images. The system includes processors and memory storing instructions. Upon receiving a text query from a user device, the system determines if the query falls into a predefined category. If it does, an option to refine the search with an image is presented. The system then returns initial results for the text query. Subsequently, if the user provides an image of an object as part of this refinement, the system generates a feature vector representing that object. This object feature vector is then compared against feature vectors of segments within the previously returned image results, yielding similarity scores. A ranked list of the initial image results is generated, prioritizing those most similar to the user's provided object image based on these similarity scores. This ranked list is then presented to the user.
2. The computing system of claim 1 , wherein each of the plurality of stored image segments corresponds to less than an entire image.
3. The computing system of claim 2 , wherein the image is at least one of an image generated by a camera of a user device, an image obtained from a memory of the user device, an image obtained from the plurality of results, or an image obtained from a storage medium that is remote from the user device.
This invention relates to a computing system for processing and analyzing images, addressing the need for flexible image sourcing in applications such as augmented reality, image recognition, or content retrieval. The system is designed to handle images from multiple sources, ensuring adaptability and robustness in various use cases. The computing system processes an image to generate a plurality of results, which may include augmented reality overlays, object recognition data, or other derived information. The image input can originate from diverse sources, including a camera of a user device, a local memory of the user device, a remote storage medium, or even from the system's own generated results. This flexibility allows the system to operate in real-time with live camera feeds, retrieve stored images for analysis, or reuse previously processed images for further refinement or comparison. By supporting multiple image sources, the system enhances usability in scenarios where images may be captured dynamically, retrieved from storage, or shared across devices. The ability to source images from remote storage or the system's own results enables advanced workflows, such as iterative analysis or collaborative applications. The invention ensures seamless integration with different image acquisition methods, improving efficiency and versatility in image-based computing tasks.
4. The computing system of claim 1 , wherein the defined category is at least one of fashion, clothing, home decor, personal, or food.
This invention relates to a computing system for categorizing and analyzing visual content, particularly images or videos, to identify and classify objects or items within predefined categories. The system addresses the challenge of efficiently organizing and retrieving visual data by automating the classification process, reducing manual effort, and improving accuracy. The computing system processes input visual content to detect and recognize objects or items present in the content. It then categorizes these detected objects into at least one of several predefined categories, including fashion, clothing, home decor, personal items, or food. The system may use machine learning models, such as convolutional neural networks, to analyze visual features and assign appropriate categories to the detected objects. Additionally, the system may include a user interface for displaying the categorized content, allowing users to filter or search based on the assigned categories. The system may also support further analysis, such as identifying subcategories within the broader categories (e.g., types of clothing like shirts, pants, or accessories) or extracting metadata related to the detected objects (e.g., color, style, or brand). This enables applications in e-commerce, content management, and personalized recommendations. The system may also integrate with external databases or APIs to enhance categorization accuracy or provide additional context.
5. A computer-implemented method, comprising: receiving, from a user device, a query; determining that the query corresponds to a defined category; presenting a visual refinement option in response to a determination that the query corresponds to the defined category; determining a first plurality of images based at least in part on the query; subsequent to receiving the query, receiving an image of an obj ect as part of the visual refinement option; comparing the image of the object with at least one image segment of each of the first plurality of images; determining, for each of the first plurality of images, a respective similarity score representative of a similarity between the image of the object and the at least one image segment of a respective one of the first plurality of images; determining a ranked list of at least a portion of the first plurality of images based at least in part on the respective similarity scores; and providing for presentation, according to the ranked list, the at least a portion of the first plurality of images.
6. The computer-implemented method of claim 5 , further comprising: processing the image to determine an object type of the object represented in the image; and wherein comparing the image includes: generating an object feature vector representative of the object; and comparing the object feature vector with a plurality of stored feature vectors corresponding to objects represented in the first plurality of images having a same object type.
This invention relates to computer vision and image processing, specifically for identifying and comparing objects within images. The problem addressed is the need for accurate and efficient object recognition and comparison in large image datasets, particularly when objects of the same type must be distinguished based on their features. The method involves processing an image to determine the type of object it represents. Once the object type is identified, the image is further processed to generate an object feature vector—a numerical representation of the object's distinguishing characteristics. This feature vector is then compared against a database of stored feature vectors corresponding to objects of the same type from previously analyzed images. The comparison helps identify similarities or differences between the object in the new image and those in the database, enabling tasks such as object matching, classification, or retrieval. The method ensures that comparisons are made only between objects of the same type, improving accuracy by reducing irrelevant matches. This approach is useful in applications like inventory management, quality control, or automated image cataloging, where precise object identification is critical. The system leverages machine learning or pattern recognition techniques to extract and compare feature vectors, enhancing the efficiency and reliability of object recognition in large-scale image processing systems.
7. The computer-implemented method of claim 5 , further comprising: detecting the object within a field of view of a camera of the user device; determining, at the user device, an object type corresponding to the object; and presenting on a display of the user device an object type identifier.
8. The computer-implemented method of claim 7 , wherein the object type identifier includes at least one of a graphical presentation corresponding to a shape of the object type, or a name of the object type.
9. The computer-implemented method of claim 7 , further comprising: enabling a selection of a second object type identifier, thereby indicating a second object type corresponding to the object.
10. The computer-implemented method of claim 9 , further comprising: generating a keyword corresponding to the object type; and including the keyword as part of the query.
11. The computer-implemented method of claim 5 , further comprising: determining a plurality of keywords corresponding to at least one of the portion of the first plurality of images, the first plurality of images, or the query; and providing for presentation and selection by a user of the user device, each of the plurality of keywords.
12. The computer-implemented method of claim 5 , wherein comparing the image of the object, further includes: generating an object feature vector representative of the object; determining an area of interest in the first plurality of images based on an object type of the object; determining the at least one image segment based at least in part on the area of interest; and comparing the object feature vector with stored feature vectors associated with the image segments.
13. The computer-implemented method of claim 12 , further comprising: determining an expected position of the object; and wherein determining the expected position is based at least in part on the object type.
14. The computer-implemented method of claim 5 , wherein the visual refinement option consists of the receiving the image of the object.
15. The computer-implemented method of claim 5 , wherein the ranked list of at least a portion of the first plurality of images is a blended result of the at least a portion of the first plurality of images.
16. A non-transitory computer-readable storage medium storing instructions that, when executed by at least one processor of a computing system, cause the computing system to at least: receive a query; determine that the query corresponds to a defined category; enable a visual refinement option on the user device; receive a streaming video from the user device as part of the visual refinement option, wherein the streaming video includes a live-feed video from a camera of the user device in real-time; process at least a portion of the streaming video to identify an object type of one or more objects represented in the streaming video; present, on a display of the user device and concurrent with a presentation of the streaming video, the object type of the one or more objects; receive from the user device a selection of the object type; determine a plurality of stored images that correspond to both the query and the selected object type; and present on the display of the user device the plurality of stored images.
17. The non-transitory computer-readable storage medium of claim 16 , wherein: the defined category is food; and the streaming video includes representations of food currently within a field of view of a camera of the user device.
18. The non-transitory computer-readable storage medium of claim 17 , wherein the query is a text based query.
19. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions further cause the computing system to at least: determine, based at least in part on the defined category, candidate images to be considered in determining the plurality of stored images.
20. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions that cause the at least one processor to determine the plurality of stored images further cause the computing system to at least: determine at least one keyword corresponding to the query, the at least one object type, or the defined category; and determine the plurality of stored images based at least in part on the at least one keyword.
This invention relates to a computer-implemented system for retrieving and displaying images based on user queries, object types, or predefined categories. The system addresses the challenge of efficiently identifying and presenting relevant images from a stored collection in response to user input. The invention involves a computing system that processes a query to determine a plurality of stored images, where the images are selected based on their relevance to the query, an object type, or a defined category. The system further refines the image selection by determining at least one keyword associated with the query, object type, or category, and then filtering or ranking the stored images based on these keywords. This keyword-based approach enhances the accuracy and relevance of the retrieved images, ensuring that the most pertinent visual content is presented to the user. The system may also include additional features such as displaying the images in a user interface, allowing user interaction, or further processing the images for analysis or annotation. The overall goal is to improve the efficiency and effectiveness of image retrieval in digital systems.
Unknown
March 9, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.